SOPRANO: Macitentan in patients with pulmonary hypertension following left ventricular assist device implantation

Abstract

Macitentan is a dual endothelin receptor antagonist (ERA) approved for treating pulmonary arterial hypertension (PAH). SOPRANO evaluated the efficacy and safety of macitentan versus placebo in pulmonary hypertension (PH) patients after left ventricular assist device (LVAD) implantation. SOPRANO was a phase 2, multicenter, double‐blind, randomized, placebo‐controlled, parallel‐group study. Patients with an LVAD implanted within the prior 90 days who had persistent PH (i.e., mean pulmonary arterial pressure ≥25 mmHg, pulmonary artery wedge pressure [PAWP] ≤18 mmHg, and pulmonary vascular resistance [PVR] >3 Wood units [WU]) were randomized (1:1) to macitentan 10 mg or placebo once daily for 12 weeks. The primary endpoint was change in PVR. Secondary endpoints included change in right‐heart catheterization hemodynamic variables, N‐terminal prohormone of brain natriuretic peptide levels, World Health Organization functional class, and safety/tolerability. Fifty‐seven patients were randomized to macitentan (n = 28) or placebo (n = 29). A statistically significant reduction in PVR from baseline to Week 12 was observed with macitentan versus placebo (placebo‐corrected geometric mean ratio, 0.74; 95% confidence interval, 0.58–0.94; p = .0158). No statistically significant differences were observed in secondary endpoints. In a post‐hoc analysis, 66.7% of patients receiving macitentan achieved PVR <3 WU versus 40.0% receiving placebo (p = .0383). Macitentan was generally well tolerated; adverse events were consistent with those in previous PAH studies with macitentan. In conclusion, macitentan showed promising tolerability and significantly reduced PVR in PH patients with persistently elevated PVR after LVAD implantation. ClinicalTrials. gov identifier: NCT02554903.

INTRODUCTION

The prevalence of heart failure (HF) in the United States is increasing. ¹ Left ventricular assist devices (LVADs) improve survival and quality of life in select patients with Stage D HF and reduced ejection fraction. ² However, persistent pulmonary hypertension (PH) remains an important risk factor for right‐heart failure while on device support, ³ and LVAD patients with right‐heart failure have an increased risk of primary graft dysfunction and death following heart transplant. ⁴ Elevated pulmonary vascular resistance (PVR) is a risk factor for acute right‐heart failure after transplantation and may impact eligibility, limit donor selection, and preclude patients from heart transplant. ² , ⁵ , ⁶ Furthermore, an increased diastolic pulmonary pressure gradient has been shown to be a strong predictor of death and HF readmission in LVAD patients, irrespective of PVR status. ⁷

Data regarding the appropriate use of pulmonary vasodilators in post‐LVAD patients are lacking, as are randomized studies of targeted PH therapies. ⁸ , ⁹ , ¹⁰ Current data are insufficient to support evidence‐based treatment decisions regarding the effectiveness of phosphodiesterase type 5 A inhibitors in lowering PVR or attenuating right ventricular (RV) failure after LVAD implantation. ¹¹ , ¹² Guidelines recommend caution when using oral or inhaled pulmonary vasodilators to treat patients with elevated PVR and clinically important PH. ¹³ There is, therefore, an unmet need for effective treatment options for these patients.

In a retrospective analysis of data from 50 patients with persistent PH after LVAD implantation treated with bosentan, an ERA, most patients experienced a significant decrease in pulmonary artery pressure and PVR, as well as improved RV function. These results indicate that ERAs may have a favorable impact on pulmonary vascular remodeling in patients with left‐heart disease. ⁸ Macitentan is a potent dual ERA that inhibits the binding of endothelin‐1 (ET‐1) to ET_A and ET_B receptors in human pulmonary arterial smooth muscle cells. ¹⁴ In the phase 3, randomized SERAPHIN study, macitentan significantly reduced morbidity and mortality in patients with pulmonary arterial hypertension (PAH) versus placebo, with manageable tolerability. ¹⁵ Macitentan is approved as treatment for patients with PAH in World Health Organization (WHO) Group 1. ¹⁴

In SOPRANO, we evaluated the efficacy and safety of macitentan versus placebo in patients with PH (defined as PVR >3 Wood units [WU]) after LVAD implantation. The primary endpoint was change in PVR from baseline to Week 12. The effects of macitentan on cardiopulmonary hemodynamics and on RV and renal function were also assessed.

METHODS

Study design

SOPRANO (ClinicalTrials. gov identifier: NCT02554903) was a phase 2, prospective, multicenter, double‐blind, randomized, placebo‐controlled, parallel‐group study (Supporting Information Figure S1). Patients were screened ≤90 days from LVAD implantation until randomization, followed by a 12‐week double‐blind treatment phase and a 30‐day safety follow‐up period following discontinuation of macitentan. The study planned to enroll and randomize 78 patients.

The study protocol was approved by an institutional review board and conducted in accordance with the ethical principles of the Declaration of Helsinki, Good Clinical Practices, regulatory requirements, and International Society for Heart and Lung Transplantation ethics. All patients provided written informed consent before participation.

Patients

Patients were aged ≥18 years, with mean pulmonary arterial pressure (mPAP) ≥25 mmHg at rest and PVR >3 WU. They were required to have PAWP ≤18 mmHg, to exclude those with insufficient left heart unloading. Patients had surgical LVAD implantation (e.g., HeartMate II or HeartWare) ≤90 days before being randomized. Randomization had to occur ≤14 days after the qualifying right‐heart catheterization (RHC). Patients had to be stabilized after removal of the pulmonary artery catheter, with no LVAD pump speed/flow rate changes, a stable dose of oral diuretics, no intravenous inotropes or vasopressors, and the ability to walk during the 48 h before screening.

Patients were excluded for documented severe obstructive lung disease (forced expiratory volume in 1 s [FEV₁]/forced vital capacity <0.7 associated with FEV₁ <50% of predicted value), moderate‐to‐severe restrictive lung disease (total lung capacity <60% of predicted value after bronchodilator administration), or pulmonary veno‐occlusive disease. Patients were also excluded if they had an estimated glomerular filtration rate (eGFR) <30 mL/min, were undergoing dialysis, or had severe hepatic impairment (Child–Pugh Class C), hemoglobin <8.5 g/dL, or Doppler mean blood pressure <65 mmHg at screening. No treatment with ERAs, phosphodiesterase type 5 A inhibitors, intravenous/subcutaneous/oral prostanoids, or guanylate cyclase stimulators was permitted ≤7 days before baseline RHC or study treatment initiation. No treatment with inhaled prostanoids or nitric oxide was permitted ≤24 h before baseline RHC or study treatment initiation. PAH‐specific therapies were not allowed as planned treatment during the study.

Treatment

Patients were randomized (1:1) to receive once‐daily oral macitentan 10 mg or matching placebo, every morning irrespective of food intake.

Endpoints and outcome measures

The primary endpoint was the PVR ratio of Week 12 to baseline, utilizing cardiac output measured by the thermodilution method and calculated programmatically. Sensitivity analyses tested the robustness of the primary efficacy data (see Methods section of Supporting Information). Secondary endpoints were changes from baseline to Week 12 in other investigator‐reported hemodynamic variables (i.e., mean right atrial pressure, mPAP, PAWP, cardiac index, total pulmonary resistance [calculated as mPAP/cardiac output], and mixed venous oxygen saturation), N‐terminal prohormone of brain natriuretic peptide (NT‐proBNP) levels, and WHO functional class (FC). In a post‐hoc analysis, achievement of a post‐baseline PVR <3 WU was evaluated in patients stratified by severity of baseline PVR elevation (≤4 vs. >4 WU) and according to each treatment group.

Safety assessments included adverse events (AEs), serious AEs (SAEs), AEs leading to treatment discontinuations, deaths, AEs of interest (anemia, hypotension, and edema/fluid retention), pregnancy, clinical laboratory tests, vital signs, body weight, and physical examination. Patients were assessed for AEs and SAEs at each visit. RHC, echocardiogram, NT‐proBNP, and WHO FC data were collected at screening and Week 12. Biomarker and eGFR data were collected at randomization and Week 12.

Statistical analyses

Primary and secondary efficacy analyses were performed on the full analysis set (all randomized patients). Safety analyses were performed on the safety analysis set (patients who received at least one dose of study treatment). A sample size of 70 evaluable patients (35 patients per treatment group) was required to achieve 80% power, to detect a difference in the PVR ratio of Week 12 to baseline of −0.28 (i.e., a geometric mean ratio [GMR] of 0.76 for macitentan vs. placebo). For the primary endpoint, PVR was calculated using the formula: (mPAP − PAWP)/cardiac output, where mPAP and PAWP were measured at end–expiration and cardiac output was measured in triplicate using the thermodilution method. Analysis of covariance (ANCOVA) was performed on the log‐transformed ratio of Week 12 to baseline PVR, with a main effect for treatment and a covariate for baseline log PVR. The GMR, with 95% confidence interval (CI), was calculated as the exponentiated value to least‐squares mean difference between treatments to present the results for PVR. This is a widely used statistical approach for addressing highly variable baseline data that has been applied in other studies evaluating the hemodynamic effects of treatment in patients with PAH. ¹⁶ , ¹⁷ , ¹⁸ Two‐sided p‐values were calculated, with p < .05 being statistically significant. In the post‐hoc analysis of patients stratified by baseline PVR elevation severity (≤4 vs. >4 WU), the proportions of patients achieving a post‐baseline PVR <3 WU were determined, and p‐values were calculated for the comparison of macitentan and placebo using Cochran–Mantel–Haenszel statistics adjusted for baseline PVR category.

Changes from baseline to Week 12 in secondary endpoint variables (hemodynamic measures, NT‐proBNP) were analyzed using an ANCOVA with a factor for treatment group (macitentan vs. placebo) and a covariate for baseline values. Geometric means and coefficients of variability were calculated. Changes from baseline in WHO FC were categorized as worsening (change >0), improving (change <0), or no change (change = 0). Changes in WHO FC were compared using a proportional odds model (ordered logistic regression) with covariates for treatment group and baseline WHO FC (I/II vs. III/IV). Treatment‐emergent AEs (TEAEs), abnormal laboratory tests, vital signs, and body weight measurements were summarized descriptively.

RESULTS

Patient disposition

Between March 28, 2016, and March 13, 2020, 115 patients were screened. The study was stopped early due to slow enrollment, after 57 of the planned 78 patients were randomized. Of the 57 patients randomized from 50 US sites (full analysis set), 28 received macitentan and 29 placebo (Figure 1).

Figure 1.

Patient disposition. Abbreviations: AE, adverse event; D/C, discontinued; HD, hemodynamics; LVAD, left ventricular assist device; RHC, right‐heart catheterization. ^aOther reasons included: not randomized within 14 days of baseline RHC (n = 1); not stable within 48 h before baseline RHC (n = 1); LVAD not implanted ≤90 days before randomization (n = 2); no written informed consent (n = 1); reason not available (n = 11). ^bOne patient experienced cerebral hemorrhage unrelated to the study drug. ^cOne patient reported acute kidney injury (unrelated) and worsening right‐heart failure (unrelated). ^dOne patient experienced hemorrhagic stroke (unrelated).

All randomized patients received at least one dose of study treatment and were included in the safety analysis set. Details of evaluable and per‐protocol sets used in sensitivity analyses are in the Supporting Information (Methods and Results sections). Forty‐eight (84.2%) patients completed 12 weeks of study treatment (24 [85.7%] in the macitentan group; 24 [82.8%] in the placebo group). Discontinuation rates were similar in both treatment arms (Figure 1).

Baseline demographics and disease characteristics

Patient demographics and disease characteristics at baseline were well balanced between treatment arms. The median age of patients was 58 years, and the majority (87.7%) were in WHO FC II/III. Baseline mPAP, PAWP, PVR, cardiac output, and other hemodynamic variables were similar in both arms (Table 1). The mean time (± standard deviation [SD]) from LVAD implantation to treatment start was 45.9 (± 23.59) days. Concomitant medications taken by patients in both arms are reported in the Supporting Information, Table S1.

Table 1.

Patient demographics and disease characteristics at baseline.

Characteristic	All patients (N = 57)	Macitentan (n = 28)	Placebo (n = 29)
Age, years	58.0 (54.0–63.0)	57.0 (53.0–62.0)	60.0 (55.0–63.0)
Female, n (%)	12 (21.1)	6 (21.4)	6 (20.7)
Race, n (%)
White	34 (59.6)	16 (57.1)	18 (62.1)
Black/African American	19 (33.3)	8 (28.6)	11 (37.9)
Asian	2 (3.5)	2 (7.1)	0
Othera	2 (3.5)	2 (7.1)	0
Select laboratory values
NT‐proBNP, pmol/L	2,045.5 (1,139.5–3,098.0)b	1,943.0 (1,257.0–3,377.0)b	2,148.0 (1,129.0–3,041.0)
Hemoglobin, g/dL	9.6 (8.7–11.0)c	9.7 (8.7–11.0)	9.5 (8.8–11.0)c
Vital signs, mean (SD)
Doppler systolic BP, mmHg	98.9 (16.6)	99.7 (18.0)	98.1 (15.7)
Doppler diastolic BP, mmHg	75.1 (12.1)	77.1 (12.6)	72.4 (11.6)
Pulse rate, bpm	83.7 (12.0)	86.9 (11.4)	80.6 (11.8)
WHO functional class, n (%)
I	4 (7.0)	2 (7.1)	2 (6.9)
II	17 (29.8)	7 (25.0)	10 (34.5)
III	33 (57.9)	17 (60.7)	16 (55.2)
IV	2 (3.5)	1 (3.6)	1 (3.4)
Hemodynamic parameters
mPAP, mmHg (end– expiratory mean line)	30.0 (27.0–33.0)	30.5 (27.5–33.5)	29.0 (27.0–33.0)
PAWP, mmHg (end– expiratory mean line)	13.0 (9.0–14.0)	12.5 (8.0–15.5)	13.0 (10.0–14.0)
PVR, WU (thermodilution)d	4.2 (3.4–5.0)	4.1 (3.6–5.0)	4.2 (3.4–4.8)
CO, L/min (thermodilution)d	4.5 (3.7–5.0)	4.5 (3.7–5.2)	4.5 (3.8–4.9)
mRAP, mmHg	11.0 (7.0–15.0)	11.0 (7.0–15.5)	11.0 (8.0–14.0)
SvO2, %	55.0 (49.0–62.1)	56.5 (47.5–62.6)	54.0 (50.0–62.0)

Note: Data are presented for the full analysis set. All data are median (IQR), unless otherwise stated.

Abbreviations: BP, blood pressure; bpm, beats per minute; CO, cardiac output; IQR, interquartile range; mPAP, mean pulmonary arterial pressure; mRAP, mean right atrial pressure; NT‐proBNP, N‐terminal prohormone of brain natriuretic peptide; PAWP, pulmonary artery wedge pressure; PVR, pulmonary vascular resistance; SD, standard deviation; SvO₂, mixed venous oxygen saturation; WHO, World Health Organization; WU, Wood unit.

^a Includes Latin American (n = 1) and unknown (n = 1).

^b Overall, 56 patients were evaluable, including 27 treated with macitentan.

^c Overall, 56 patients were evaluable, including 28 receiving placebo.

^d Overall, 55 patients were evaluable, including 27 treated with macitentan and 28 receiving placebo.

Reduction in PVR

The median (interquartile range) PVR for all enrolled patients was 4.2 (3.4–5.0) WU and was similar in both treatment arms. Mean (SD) PVR reduced from 4.32 (0.93) at baseline to 2.51 (1.06) at Week 12 in the macitentan arm, and from 4.31 (1.28) to 3.26 (1.34) in the placebo arm. A statistically significant greater reduction in PVR at Week 12 was observed with macitentan versus placebo (mean change in ratio of Week 12 to baseline PVR [SD]: 0.59 [0.23] vs. 0.76 [0.26]); the placebo‐adjusted GMR was 0.74 (95% CI 0.58, 0.94; p = .0158; Table 2), corresponding to a statistically significant 26% reduction in PVR with macitentan. Sensitivity analyses confirmed the consistency of the effect on PVR (Supporting Information, Table S2).

Table 2.

PVR using thermodilution cardiac output (primary analysis in the full analysis set).

Parameter	Macitentan (n = 27)	Placebo (n = 28)
Mean (SD) PVR, WUa
Baseline	4.32 (0.93)	4.31 (1.28)
Week 12	2.51 (1.06)	3.26 (1.34)
Ratio (Week 12:baseline)	0.59 (0.23)	0.76 (0.26)
GMR: macitentan vs. placebo (95% CI)	0.74 (0.58, 0.94)
p‐value	0.0158

Abbreviations: CI, confidence interval; CO, cardiac output; GMR, geometric mean ratio; mPAP, mean pulmonary arterial pressure; PAWP, pulmonary artery wedge pressure; PVR, pulmonary vascular resistance; SD, standard deviation; WU, Wood unit.

^a PVR calculated by programming using the formula: (mPAP − PAWP)/CO; where mPAP and PAWP were measured at end–expiration and CO was measured in triplicate using the thermodilution method. Imputations of missing values were performed for all PVR analyses, using the last observation carried forward (LOCF) method.

Proportion of patients achieving PVR <3 WU

Overall, 24 patients in the macitentan group and 25 in the placebo group were evaluable for analysis. Of these, 10 (41.7%) and 11 (44.0%) patients, respectively, had a baseline PVR ≤4 WU; 14 (58.3%) and 14 (56.0%), respectively, had a baseline PVR >4 WU.

A higher proportion of patients receiving macitentan achieved a Week 12 PVR <3 WU versus those receiving placebo (16 [66.7%] vs. 10 [40.0%] patients; p = .0383; Figure 2). Among patients with baseline PVR ≤4 WU, nine (90.0%) and seven (63.6%) patients receiving macitentan and placebo, respectively, achieved a Week 12 PVR <3 WU. In patients with a higher baseline PVR (>4 WU), seven (50.0%) macitentan and three (21.4%) placebo patients achieved a Week 12 PVR <3 WU.

Figure 2.

Post‐hoc analysis of the proportion of patients achieving a PVR <3 WU. Abbreviations: PVR, pulmonary vascular resistance; WU, Wood unit. Data are presented only for patients with both baseline and post‐baseline values. ^aThe p‐value was calculated based on Cochran–Mantel–Haenszel statistics by adjusting for baseline PVR category.

Other hemodynamic parameters, NT‐proBNP, WHO FC, and exploratory analyses

There were no statistically significant differences in any predefined secondary hemodynamic endpoints with macitentan versus placebo (Table 3). The mean PA pressure fell in both groups by on average 3–4 mmHg (31.40 ± 5.09 to 27.56 ± 9.24 with macitentan, 31.56 ± 6.02 to 27.70 ± 7.89 with placebo p = NS for change between groups). Cardiac output and PAWP were numerically higher at 12 weeks in the macitentan group, but neither achieved statistically significant change from baseline. The PAWP at 12 weeks was 15.6 ± 7.2 mmHg in the macitentan group and 13 ± 5.7 mmHg in the placebo group (LSM difference between groups from baseline to 12 weeks 2.35 mmHg (95% CI −1.19, 5.89), p = .1873).

Table 3.

Changes from baseline to Week 12 in secondary and selected exploratory efficacy endpoints.

Parameter	Macitentan (n = 25)		Placebo (n = 25)		Difference of LSM (95% CI)	p‐value
	Baseline	Week 12	Baseline	Week 12
Mean (SD) hemodynamic parameters
mPAP (end–expiratory), mmHg	31.40 (5.09)	27.56 (9.24)	31.56 (6.02)	27.70 (7.89)	−0.31 (−5.01, 4.40)	0.8967
mRAP, mmHg	10.80 (5.01)	11.60 (6.52)	10.30 (4.06)	8.80 (4.43)	2.14 (−0.57, 4.84)	0.1181
PAWP, mmHg	11.70 (4.39)	15.60 (7.17)	12.00 (3.79)	13.00 (5.66)	2.35 (−1.19, 5.89)	0.1873
Cardiac output, L/min	4.60 (0.99)	4.95 (1.04)	4.52 (0.77)	4.70 (1.15)	0.21 (−0.33, 0.74)	0.4405
Cardiac index, L/min/m2 a	2.35 (0.55)	2.44 (0.42)	2.23 (0.38)	2.28 (0.49)	0.12 (−0.11, 0.35)	0.3032
TPR, mmHg·min/La	7.11 (1.53)	5.76 (2.04)	7.12 (1.50)	6.12 (2.08)	−0.44 (−1.63, 0.74)	0.4562
SvO2, %	54.20 (12.87)	58.46 (10.46)	54.60 (8.65)	58.70 (5.36)	0.43 (−4.31, 5.17)	0.8563
TPG, mmHg	19.72 (3.76)	11.92 (4.91)	19.60 (6.37)	14.70 (5.28)	−2.70 (−5.40, −0.001)	0.0499
LVAD speed, rpm (overall and by pump type [below])	5,003.2 (2661.90)	5,021.6 (2643.66)	4,923.2 (2436.98)	4,926.4 (2438.99)	38.46 (−42.58, 119.50)	0.3443
HeartMate IIb	9,133.3 (163.30)	9,100.0 (414.73)	9,000.0 (244.95)	9,040.0 (167.33)	139.97 (−257.66, 537.60)	0.4221
HeartWarec	2,665.0 (120.64)	2,695.0 (100.23)	2,661.8 (155.55)	2,678.2 (116.09)	13.19 (−51.86, 78.25)	0.6759
LVAD power, wattsd	4.34 (0.69)	4.25 (0.81)	4.40 (0.85)	4.54 (1.21)	−0.25 (−0.85, 0.36)	0.4145
LVAD flow (L/min)e	4.39 (0.87)	4.67 (0.93)	4.62 (1.20)	4.38 (1.05)	0.46 (−0.004, 0.930)	0.0521
PAPi‐proga	3.52 (3.31)	3.74 (5.37)	2.97 (1.96)	3.27 (2.27)	0.29 (−2.05, 2.62)	0.8054
PAPP (end–expiratory mean line)	34.0 (16.69)	40.1 (33.59)	32.6 (17.82)	34.9 (16.57)	3.38 (−9.36, 16.13)	0.5955
PAPP (mean of the wave)	38.7 (29.24)	42.3 (39.32)	32.7 (18.44)	35.6 (17.10)	2.09 (−12.93, 17.11)	0.7807
RVSWi (g‐m/m2)‐proga	8.08 (3.18)	6.39 (2.55)	7.87 (3.62)	7.55 (2.64)	−1.06 (−2.40, 0.28)	0.1180
Stroke volume	54.4 (12.64)	71.1 (59.05)	57.7 (17.69)	62.3 (19.78)	11.23 (−13.47, 35.92)	0.3644
	Baseline	Week 12	Baseline	Week 12	GMR (95% CI)	p ‐value
Mean (SD) NT‐proBNP, pg/mLf	2,856.4 (2,208.79)	2,435.4 (2,620.75)	2,758.9 (2,746.12)	1,559.1 (1,052.68)	1.11 (0.75, 1.64)	0.5990
Median (IQR) NT‐proBNP, pg/mLf	2,218.0 (1,330.0–3,861.0)	1,098.0 (687.0–2,788.0)	2,023.5 (964.0–3,098.0)	1,400.0 (709.5–2,404.5)	NA	NA
	Baseline	Week 12	Baseline	Week 12	Odds ratio (95% CI)	p ‐value
WHO functional class, n (%)g
I/II	9 (32.1)	18 (64.3)	12 (41.4)	17 (58.6)	1.33 (0.42, 4.26)h	0.6280
III/IV	18 (64.3)	7 (25.0)	17 (58.6)	8 (27.6)
Missing	1 (3.6)	3 (10.7)	0	4 (13.8)

Note: Data are presented for the full analysis set, with analyses based on patients with both baseline and post‐baseline values (i.e., 25 patients in each arm, unless otherwise stated); no imputation of missing values was performed. Changes from baseline to Week 12 for each parameter were analyzed using an analysis of covariance with a factor for treatment group (macitentan vs. placebo) and a covariate for the baseline value. The difference of least square mean was calculated as described in the methods.

Abbreviations: CI, confidence interval; GMR, geometric mean ratio; IQR, interquartile range; LSM, least‐square mean; LVAD, left ventricular assist device; mPAP, mean pulmonary arterial pressure; mRAP, mean right atrial pressure; NA, not applicable; NT‐proBNP, N‐terminal prohormone of brain natriuretic peptide; PAPi, pulmonary artery pulsatility index; PAPP, pulmonary artery pulse pressure; PAWP, pulmonary artery wedge pressure; RVSWi, right ventricular stroke work index; SD, standard deviation; SvO2, mixed venous oxygen saturation; TPG, transpulmonary gradient; TPR, total pulmonary resistance; WHO, World Health Organization.

^a The number of evaluable patients was 24 for macitentan and 25 for placebo.

^b LVAD speed between 6,000 and 10,000 rpm was classified as Heartmate II. The number of evaluable patients was 5 for macitentan and 5 for placebo.

^c LVAD speed between 2,200 and 3,200 rpm was classified as HeartWare. The number of evaluable patients was 12 for macitentan and 11 for placebo.

^d The number of evaluable patients was 22 for macitentan and 22 for placebo.

^e The number of evaluable patients was 23 for macitentan and 22 for placebo.

^f The number of evaluable patients was 22 for macitentan and 24 for placebo.

^g The number of evaluable patients was 28 for macitentan and 29 for placebo.

^h The number of evaluable patients was 25 for macitentan and 25 for placebo.

There was a nonsignificant trend toward higher calculated LVAD flow in the macitentan group. There was no significant change in NT‐proBNP in the macitentan group versus the placebo group. Mean values numerically appear lower at 12 weeks in the placebo group, which may reflect a skewed distribution with outliers; however, median values were similar. A statistically significant reduction in the transpulmonary gradient (TPG) from baseline to Week 12 was observed with macitentan versus placebo. Change in pulmonary artery pulse pressure and stroke volume from baseline to 12 weeks between the groups was also not significant.

For other exploratory endpoints (Supporting Information, Supplementary results), there were no differences between treatment groups, except for changes in ET‐1 and neutrophil gelatinase‐associated lipocalin (NGAL). There were no clinically notable differences in vital signs during the study (mean [SD] changes from baseline to Week 12 in Doppler systolic/diastolic blood pressure were −4.1 [23.5]/ − 5.8 [15.4] mmHg in the macitentan and 9.4 [21.4]/8.7 [15.1] mmHg in the placebo group; Supporting Information, Table S3). Analysis of echocardiographic variables was limited by missing data, primarily because the studies were technically challenging to conduct and inadequate/low quality views prevented capture of key data by the core laboratory supporting echocardiographic analyses.

Safety and tolerability

Overall, TEAE rates (see Table 4) were similar between treatment arms, with 26 (92.9%) patients receiving macitentan and 25 (86.2%) receiving placebo experiencing at least one TEAE. Two patients experienced gastrointestinal hemorrhage (macitentan, n = 0; placebo, n = 2). The only other gastrointestinal bleeding event reported was gingival bleeding, in one patient in the placebo group. Among other TEAEs previously reported with macitentan, ¹⁵ headache occurred in one (3.6%) patient receiving macitentan and two (6.9%) receiving placebo; nasopharyngitis was not reported; and anemia, an event of interest, occurred as an SAE in one (3.4%) patient receiving placebo. Other events of interest included oedema/fluid retention occurring in four (14.3%) patients receiving macitentan and one (3.4%) receiving placebo, and hypotension occurring in three (10.7%) patients receiving macitentan and one (3.4%) receiving placebo. The overall incidence of SAEs was similar with macitentan and placebo. The most common SAE occurring in two or more patients in either arm was anticoagulation drug levels below the therapeutic level.

Table 4.

Most frequent TEAEs (reported in ≥10% of either treatment group) and SAEs (reported in ≥2 patients in either treatment group).

Preferred term	Macitentan (n = 28)	Placebo (n = 29)
Any TEAE, n (%)	26 (92.9)	25 (86.2)
Dizziness	4 (14.3)	7 (24.1)
Anticoagulation drug below therapeutic level	4 (14.3)	4 (13.8)
NT‐proBNP increased	4 (14.3)	4 (13.8)
Vomiting	2 (7.1)	4 (13.8)
Dyspnea	4 (14.3)	1 (3.4)
Fatigue	4 (14.3)	1 (3.4)
Edema peripheral	4 (14.3)	1 (3.4)
Ventricular tachycardia	1 (3.6)	4 (13.8)
Acute kidney injury	3 (10.7)	1 (3.4)
Hypotension	3 (10.7)	1 (3.4)
Rash	1 (3.6)	3 (10.3)
Complication associated with device	0	3 (10.3)
Cystatin C increased	3 (10.7)	0
Dehydration	0	3 (10.3)
Fall	0	3 (10.3)
Hyperkalemia	3 (10.7)	0
Any SAE, n (%)	16 (57.1)	16 (55.2)
Anticoagulation drug below therapeutic level	3 (10.7)	4 (13.8)
Acute kidney injury	2 (7.1)	1 (3.4)
Ventricular tachycardia	0	3 (10.3)
Pleural effusion	2 (7.1)	0
Right ventricular failure	2 (7.1)	0
Septic shock	2 (7.1)	0
Complication associated with device	0	2 (6.9)
Dehydration	0	2 (6.9)
Gastrointestinal hemorrhage	0	2 (6.9)

Note: Data are presented for the safety analysis set (i.e. all patients who received ≥1 dose of study treatment).

Abbreviations: NT‐proBNP, N‐terminal prohormone of brain natriuretic peptide; SAE, serious adverse event; TEAE, treatment‐emergent adverse event.

There were three deaths due to TEAEs: one macitentan patient died from cerebral hemorrhage; one placebo patient died from hemolysis and hemorrhagic stroke, and another from device‐related complications. No deaths were considered related to study treatment.

DISCUSSION

The SOPRANO study evaluated an early post‐LVAD population. Early, intermediate‐term, and late RV failure is a serious complication of LVAD therapy. ³ Intrinsic RV function, severity of tricuspid regurgitation, volume status, and extent of RV afterload elevation are key determinants of RV failure in this setting. ¹⁹ RV afterload includes components related to residual elevation of left heart filling pressures, which should be minimized with optimization of pump speed and diuretic dosing. The component of RV afterload related to residual elevation of PVR is clinically important because elevated PVR is a significant risk factor for death in patients with HF, including heart transplant recipients, ⁶ , ²⁰ , ²¹ , ²² and near‐normal PVR in heart transplant candidates allows for a broader range of suitable donors ⁵ , ²³ , ²⁴ ; furthermore, residual elevation of PVR also contributes to RV failure following LVAD implantation. ³ , ¹⁹

Combined pre‐ and post‐capillary PH in HF with reduced ejection fraction reflects a combination of pulmonary vasoconstriction and variable degrees of pulmonary vascular remodeling that may indicate the severity and duration of left heart filling pressure elevation. LVADs are sometimes implanted in patients with a level of PVR elevation (e.g. >5 WU) that precludes heart transplant; these patients may become transplant candidates if there is sufficient PVR decline following LVAD implantation. ²³ , ²⁵ Some patients have persistent PH after the LVAD has been implanted, which contributes to risk of RV failure, prompting treatment with vasodilators, despite limited evidence for safety and efficacy in this setting ³ and guidelines recommending caution with this treatment approach. ¹³ Nevertheless, the ERA bosentan showed promising efficacy in an uncontrolled, open‐label study, significantly reducing mPAP and PVR in patients with PH after LVAD. ⁸ Macitentan is an ERA designed to improve efficacy and safety compared with other ERAs. ²⁶ , ²⁷ SOPRANO is the first prospective, randomized controlled trial evaluating the efficacy and safety of a pulmonary vasodilator (macitentan) in the setting of post‐LVAD PH with persistently elevated PVR.

Treatment with oral macitentan 10 mg for 12 weeks achieved a 26% reduction in PVR compared with placebo in this early post‐LVAD population. A statistically significant decrease in the TPG was observed with macitentan versus placebo (p = .0499). The greater reduction in PVR in the macitentan group occurred without a greater reduction in mPAP than in the placebo group. This may partly reflect the trend toward slightly higher cardiac index in the macitentan group, combined with some expansion of plasma volume, a known effect of ERAs, with a non‐statistically significant trend towards a rise in PAWP with macitentan versus placebo. Fortunately, left heart filling pressures in LVAD recipients can be modulated by pump speed adjustment. Follow‐up hemodynamics and pump speed adjustment after initiation of pulmonary vasodilators to maintain optimal LV unloading, combined with optimization of diuretic dosing, may help to leverage the benefits of the achieved reduction in PVR by further lowering PAWP and thereby PA pressure.

The trend toward decline in mPAP and PVR in the placebo group might represent some time‐dependent reverse remodeling. However, the greater reduction in PVR with macitentan could provide benefit at time of transplant, since the immediate post‐transplant PVR remains important to post‐transplant pulmonary artery pressure and donor RV performance.

No other secondary or exploratory endpoints were statistically significant, except for changes in select biomarkers, including ET‐1 and NGAL. The analysis of secondary parameters (e.g., RV function) was severely constrained by missing echocardiography data. The observed changes in ET‐1 were expected, given that macitentan provides sustained blockade of both ET_A and ET_B receptors. ¹⁴ Reduction in NGAL may be a marker of improvement in HF severity ²⁸ ; however, given the exploratory nature of the biomarker analyses and the small sample size, these findings are hypothesis‐generating only. Macitentan was generally well tolerated, with an AE profile consistent with the underlying diseases and other macitentan studies in patients with PAH. Current results provide evidence that suggests there is no harm in using an ERA in this patient population; however, findings should be further explored in long‐term analyses.

Importantly, in the post‐hoc analysis of SOPRANO, it was shown that 66.7% of patients on macitentan achieved a post‐baseline PVR <3 WU (vs. 40.0% on placebo; p = .0383). Furthermore, in the subgroup with a baseline PVR >4 WU, approximately 2.5 times as many patients on macitentan achieved a PVR <3 WU versus those on placebo, although the small sample size precludes formal statistical comparison. This suggests that a post‐LVAD PVR approaching normal can be achieved with macitentan treatment. The importance of modest elevation in post‐LVAD PVR reflects the hyperbolic relationship between PVR and pulmonary artery capacitance. ²⁹ As the inflection point of this relationship is around 3 WU, achieving a PVR <3 WU is physiologically meaningful in the LVAD setting. Even modest improvement in PVR in this range has an amplified effect on pulmonary artery capacitance, which enhances the ability of a failing RV post LVAD to eject into the pulmonary artery. Accordingly, we anticipate that the PVR improvements seen with macitentan in SOPRANO may have clinically important implications for post‐LVAD RV function. However, the trend toward higher PAWP suggests need to optimize LVAD pump speed. In addition, for those patients post LVAD who do not have sufficient time‐dependent decline in PVR to be in optimal condition for heart transplant, we have shown that macitentan can result in greater decline in PVR than LVAD implant alone.

STUDY LIMITATIONS

Our study has several limitations. Early termination due to slow enrollment resulted in the planned sample size not being reached, and this reduced the study's power to detect statistically significant differences between treatment groups. Due to the severity of disease in this study population, there were considerable numbers of non‐evaluable patients (>10%) and protocol deviations. The short study duration (12 weeks), small sample size, and missing echocardiography data precluded meaningful analysis of effects on RV function and other clinical outcomes beyond safety. Availability of echocardiography data from more patients would have been particularly useful, as a demonstration of improvement in this parameter would have strengthened our findings. We did not include a functional capacity test such as the 6‐min Walk Test, or capture parameters which cover feel/function that may be meaningful to patients. Our finding of a greater reduction in PVR with macitentan versus placebo suggests that macitentan could reduce the risk of requiring an RVAD or other RV support post‐LVAD; however, the study design that excluded the very early postoperative period did not allow us to evaluate this outcome. Although we anticipate that the type of LVAD used would not affect the results of the study (as per the inclusion criteria, patients were required to have PAWP ≤18 mmHg, to exclude patients with insufficient left heart unloading), we acknowledge a lack of patients with contemporary HeartMate 3 LVADs as the study predated approval of this device.

CONCLUSIONS

Macitentan was generally well tolerated and effectively reduced PVR in patients with PH and elevated PVR after LVAD implantation. A statistically significant 26% reduction in PVR occurred with macitentan versus placebo (p = .0158), and more patients receiving macitentan achieved PVR <3 WU compared with placebo (66.7% vs. 40.0%; p = .0383). Macitentan was also associated with a statistically significant reduction in TPG versus placebo (p = 0.0499). Although we recognize that our study is only hypothesis‐generating, PVR reduction with macitentan in LVAD recipients could have a favorable impact on post‐LVAD right‐heart failure and on transplant eligibility and donor selection. More research is needed to understand the clinical relevance of reducing PVR following LVAD implantation and the potential clinical benefits of macitentan in this setting.

AUTHOR CONTRIBUTIONS

Robert P. Frantz, Mark A. Rocco, Mona Selej, Carol Zhao, and J. Eduardo Rame were involved in study concept and design. Robert P. Frantz, Gregory Ewald, Veronica Franco, Antoine Hage, Evelyn M. Horn, Michael A. Mathier, Stacy Mandras, Myung H. Park, I‐wen Wang, Carol Zhao, J. Eduardo Rame were involved in data acquisition, analysis, and interpretation. Carol Zhao performed the statistical analyses. Gabriela Gomez Rendon was involved in data interpretation, critical appraisal and writing of the manuscript and approved the final submission. All authors contributed to the critical appraisal and writing of the manuscript and approved the final submission.

CONFLICT OF INTEREST STATEMENT

Robert P. Frantz is receiving grants and research support from United Therapeutics, Medtronic, and Gossamer Bio, and is scientific medical advisor to Altavant, ShouTi, Liquidia Corporation, Merck, Tenax Therapeutics, and Janssen Pharmaceutical Companies of Johnson & Johnson. His institution has received funding from Bayer and Gossamer Bio. Shashank S. Desai has served on the speakers’ bureau for Abbott. Gregory Ewald is on the speakers’ bureau and is a consultant with Abbott. Veronica Franco's institution receives research support from Acceleron/Merck, Gossamer, Janssen, United Therapeutics, Aerovate Therapeutics, Respira, and Cereno Scientific. Antoine Hage is receiving grants and research support from United Therapeutics, Lung LLC, Arena, Reata, and Bayer, and is a stockholder in Johnson & Johnson and Pfizer. Evelyn M. Horn's institution has received funding from Gossamer, Acceleron/Merck, Abbott, and Cereno Scientific. Stacy Mandras is a speaker and consultant for United Therapeutics and Bayer Pharmaceuticals. Myung H. Park is a scientific medical advisor with AstraZeneca. Ashwin K. Ravichandran is consulting/speaking for Abbott, Medtronic and speaking for United Therapeutics, Janssen, and Bayer; he personally receives no grants from these companies, but his institution does. Ronald Zolty is a consultant for Janssen Pharmaceutical Companies of Johnson & Johnson, Bayer, United Therapeutics, and Alnylam. Mark A. Rocco, Mona Selej, and Carol Zhao were employees of Actelion Pharmaceuticals US, Inc., South San Francisco, CA (at the time of manuscript development) and are current stockholders of Johnson & Johnson. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.

ETHICAL STATEMENT

The study protocol was approved by an institutional review board and conducted in accordance with the ethical principles of the Declaration of Helsinki, Good Clinical Practices, regulatory requirements, and International Society for Heart and Lung Transplantation ethics. All patients provided written informed consent before participation.

GUARANTOR

Dr Robert P. Frantz, Mayo Clinic, Rochester, MN, USA.

Supporting information

Supporting information.

Frantz RP, Desai SS, Ewald G, Franco V, Hage A, Horn EM, LaRue SJ, Mathier MA, Mandras S, Park MH, Ravichandran AK, Schilling JD, Wang I‐w, Zolty R, Rendon GG, Rocco MA, Selej M, Zhao C, Rame JE. SOPRANO: Macitentan in patients with pulmonary hypertension following left ventricular assist device implantation. Pulm Circ. 2024;14:e12446. 10.1002/pul2.12446